6e403221 |
1 | /* -*-c-*- |
2 | * |
a10122de |
3 | * $Id: anag.c,v 1.5 2002/08/11 12:58:09 mdw Exp $ |
6e403221 |
4 | * |
5 | * Main driver for anag |
6 | * |
7 | * (c) 2001 Mark Wooding |
8 | */ |
9 | |
10 | /*----- Licensing notice --------------------------------------------------* |
11 | * |
12 | * This file is part of Anag: a simple wordgame helper. |
13 | * |
14 | * Anag is free software; you can redistribute it and/or modify |
15 | * it under the terms of the GNU General Public License as published by |
16 | * the Free Software Foundation; either version 2 of the License, or |
17 | * (at your option) any later version. |
18 | * |
19 | * Anag is distributed in the hope that it will be useful, |
20 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
21 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
22 | * GNU General Public License for more details. |
23 | * |
24 | * You should have received a copy of the GNU General Public License |
25 | * along with Anag; if not, write to the Free Software Foundation, |
26 | * Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. |
27 | */ |
28 | |
29 | /*----- Revision history --------------------------------------------------* |
30 | * |
31 | * $Log: anag.c,v $ |
a10122de |
32 | * Revision 1.5 2002/08/11 12:58:09 mdw |
33 | * Added support for regular expression matching, if supported by the C |
34 | * library. |
35 | * |
1d2d1062 |
36 | * Revision 1.4 2001/02/19 19:18:50 mdw |
37 | * Minor big fixes to parser. |
38 | * |
2668675c |
39 | * Revision 1.3 2001/02/16 21:45:19 mdw |
40 | * Be more helpful. Improve full help message. Special-case error for |
41 | * empty command strings. |
42 | * |
60dffc01 |
43 | * Revision 1.2 2001/02/07 09:09:11 mdw |
44 | * Fix spurious error when `-file' is used. |
45 | * |
6e403221 |
46 | * Revision 1.1 2001/02/04 17:14:42 mdw |
47 | * Initial checkin |
48 | * |
49 | */ |
50 | |
51 | /*----- Header files ------------------------------------------------------*/ |
52 | |
53 | #include "anag.h" |
54 | |
55 | /*----- Static variables --------------------------------------------------*/ |
56 | |
57 | static const char *file = DICTIONARY; |
58 | |
59 | /*----- Help text functions -----------------------------------------------*/ |
60 | |
61 | static void usage(FILE *fp) |
62 | { |
63 | pquis(fp, "Usage: $ [-f file] expression\n"); |
64 | } |
65 | |
66 | static void version(FILE *fp) |
67 | { |
68 | pquis(fp, "$, version " VERSION "\n"); |
69 | } |
70 | |
71 | static void help(FILE *fp) |
72 | { |
73 | version(fp); |
74 | fputc('\n', fp); |
75 | usage(fp); |
76 | fputs("\n\ |
77 | Searches a wordlist, printing all of the words which match an expression.\n\ |
2668675c |
78 | \n\ |
79 | Options supported are:\n\ |
80 | \n\ |
81 | -h, --help display this help text\n\ |
82 | -v, --version display the program's version number\n\ |
83 | -u, --usage display a very brief usage message\n\ |
84 | -f, --file FILE read wordlist from FILE, not `" DICTIONARY "'\n\ |
85 | \n\ |
6e403221 |
86 | The basic tests in the expression are:\n\ |
87 | \n\ |
88 | -anagram WORD matches a full-length anagram\n\ |
89 | -subgram WORD matches words which only use letters in WORD\n\ |
90 | -wildcard PATTERN matches with wildcards `*' and `?'\n\ |
91 | -trackword WORD matches words which can be found in a trackword\n\ |
a10122de |
92 | " |
93 | #ifdef HAVE_REGCOMP |
94 | "\ |
95 | -regexp REGEXP matches with an (extended) regular expression\n\ |
96 | " |
97 | #endif |
98 | "\ |
6e403221 |
99 | \n\ |
100 | These simple tests can be combined using the operators `-a', `-o' and `-n'\n\ |
101 | (for `and', `or' and `not'; they may also be written `&', `|' and `!' if\n\ |
102 | you like), and grouped using parentheses `(' and `)'.\n\ |
103 | ", fp); |
104 | } |
105 | |
106 | /*----- The options parser ------------------------------------------------*/ |
107 | |
108 | /* --- Options table structure --- */ |
109 | |
110 | struct opt { |
111 | const char *name; |
112 | unsigned nargs; |
113 | unsigned f; |
114 | unsigned tag; |
115 | }; |
116 | |
117 | enum { |
118 | O_HELP, O_VERSION, O_USAGE, |
119 | O_FILE, |
120 | O_AND, O_OR, O_NOT, O_LPAREN, O_RPAREN, |
a10122de |
121 | O_ANAG, O_SUBG, O_WILD, O_TRACK, O_REGEXP, |
6e403221 |
122 | O_EOF |
123 | }; |
124 | |
125 | #define OF_SHORT 1u |
126 | |
127 | static const struct opt opttab[] = { |
128 | |
129 | /* --- Options -- don't form part of the language --- */ |
130 | |
131 | { "help", 0, OF_SHORT, O_HELP }, |
132 | { "version", 0, OF_SHORT, O_VERSION }, |
133 | { "usage", 0, OF_SHORT, O_USAGE }, |
134 | { "file", 1, OF_SHORT, O_FILE }, |
135 | |
136 | /* --- Operators -- provide the basic structure of the language --- * |
137 | * |
138 | * These are also given magical names by the parser. |
139 | */ |
140 | |
141 | { "and", 0, OF_SHORT, O_AND }, |
142 | { "or", 0, OF_SHORT, O_OR }, |
143 | { "not", 0, OF_SHORT, O_NOT }, |
144 | |
145 | /* --- Actual matching oeprations -- do something useful --- */ |
146 | |
147 | { "anagram", 1, 0, O_ANAG }, |
148 | { "subgram", 1, 0, O_SUBG }, |
149 | { "wildcard", 1, 0, O_WILD }, |
150 | { "trackword", 1, 0, O_TRACK }, |
a10122de |
151 | #ifdef HAVE_REGCOMP |
152 | { "regexp", 1, 0, O_REGEXP }, |
153 | #endif |
6e403221 |
154 | |
155 | /* --- End marker --- */ |
156 | |
157 | { 0, 0, 0, 0 } |
158 | }; |
159 | |
160 | static int ac; |
161 | static const char *const *av; |
162 | static int ai; |
163 | |
164 | /* --- @nextopt@ --- * |
165 | * |
166 | * Arguments: @const char ***arg@ = where to store the arg pointer |
167 | * |
168 | * Returns: The tag of the next option. |
169 | * |
170 | * Use: Scans the next option off the command line. If the option |
171 | * doesn't form part of the language, it's processed internally, |
172 | * and you'll never see it from here. On exit, the @arg@ |
173 | * pointer is set to contain the address of the option scanned, |
174 | * followed by its arguments if any. You're expected to know |
175 | * how many arguments there are for your option. |
176 | */ |
177 | |
178 | static unsigned nextopt(const char *const **arg) |
179 | { |
180 | for (;;) { |
181 | const struct opt *o, *oo; |
182 | size_t sz; |
183 | const char *p; |
184 | |
185 | /* --- Pick the next option off the front --- */ |
186 | |
187 | *arg = av + ai; |
188 | if (ai >= ac) |
189 | return (O_EOF); |
190 | p = av[ai++]; |
191 | |
192 | /* --- Cope with various forms of magic --- */ |
193 | |
194 | if (p[0] != '-') { |
195 | if (!p[1]) switch (*p) { |
196 | case '&': return (O_AND); |
197 | case '|': return (O_OR); |
198 | case '!': return (O_NOT); |
199 | case '(': return (O_LPAREN); |
200 | case ')': return (O_RPAREN); |
201 | } |
202 | goto bad; |
203 | } |
204 | |
205 | /* --- Now cope with other sorts of weirdies --- * |
206 | * |
207 | * By the end of this, a leading `-' or `--' will have been stripped. |
208 | */ |
209 | |
210 | p++; |
211 | if (!*p) |
212 | goto bad; |
213 | if (*p == '-') |
214 | p++; |
215 | if (!*p) { |
216 | if (ai < ac) |
217 | die("syntax error near `--': rubbish at end of line"); |
218 | return (O_EOF); |
219 | } |
220 | |
221 | /* --- Now look the word up in my table --- */ |
222 | |
223 | sz = strlen(p); |
224 | oo = 0; |
225 | for (o = opttab; o->name; o++) { |
226 | if (strncmp(p, o->name, sz) == 0) { |
227 | if (strlen(o->name) == sz || ((o->f & OF_SHORT) && sz == 1)) { |
228 | oo = o; |
229 | break; |
230 | } |
231 | if (oo) { |
232 | die("ambiguous option name `-%s' (could match `-%s' or `-%s')", |
233 | p, oo->name, o->name); |
234 | } |
235 | oo = o; |
236 | } |
237 | } |
238 | if (!oo) |
239 | die("unrecognized option name `-%s'", p); |
240 | |
241 | /* --- Sort out the arguments --- */ |
242 | |
243 | if (ai + oo->nargs > ac) |
244 | die("too few arguments for `-%s' (need %u)", oo->name, oo->nargs); |
245 | ai += oo->nargs; |
246 | |
247 | /* --- Now process the option --- */ |
248 | |
249 | switch (oo->tag) { |
250 | case O_HELP: |
251 | help(stdout); |
252 | exit(0); |
253 | case O_VERSION: |
254 | version(stdout); |
255 | exit(0); |
256 | case O_USAGE: |
257 | usage(stdout); |
258 | exit(0); |
259 | case O_FILE: |
260 | file = (*arg)[1]; |
261 | break; |
262 | default: |
263 | return (oo->tag); |
264 | } |
60dffc01 |
265 | continue; |
6e403221 |
266 | bad: |
267 | die("syntax error near `%s': unknown token type", av[ai - 1]); |
268 | } |
269 | } |
270 | |
271 | /*----- Node types for operators ------------------------------------------*/ |
272 | |
273 | /* --- Node structures --- */ |
274 | |
275 | typedef struct node_bin { |
276 | node n; |
277 | node *left; |
278 | node *right; |
279 | } node_bin; |
280 | |
281 | typedef struct node_un { |
282 | node n; |
283 | node *arg; |
284 | } node_un; |
285 | |
286 | /* --- Node functions --- */ |
287 | |
288 | static int n_or(node *nn, const char *p, size_t sz) |
289 | { |
290 | node_bin *n = (node_bin *)nn; |
291 | return (n->left->func(n->left, p, sz) || n->right->func(n->right, p, sz)); |
292 | } |
293 | |
294 | static int n_and(node *nn, const char *p, size_t sz) |
295 | { |
296 | node_bin *n = (node_bin *)nn; |
297 | return (n->left->func(n->left, p, sz) && n->right->func(n->right, p, sz)); |
298 | } |
299 | |
300 | static int n_not(node *nn, const char *p, size_t sz) |
301 | { |
302 | node_un *n = (node_un *)nn; |
303 | return (!n->arg->func(n->arg, p, sz)); |
304 | } |
305 | |
306 | /*----- Parser for the expression syntax ----------------------------------*/ |
307 | |
308 | /* --- A parser context --- */ |
309 | |
310 | typedef struct p_ctx { |
311 | unsigned t; |
312 | const char *const *a; |
313 | } p_ctx; |
314 | |
315 | /* --- Parser structure --- * |
316 | * |
317 | * This is a simple recursive descent parser. The context retains |
318 | * information about the current token. Each function is passed the address |
319 | * of a node pointer to fill in. This simplifies the binary operator code |
320 | * somewhat, relative to returning pointers to node trees. |
321 | */ |
322 | |
323 | static void p_expr(p_ctx *p, node **/*nn*/); |
324 | |
325 | static void p_next(p_ctx *p) |
326 | { |
327 | static const char *const eof[] = { "<end>", 0 }; |
328 | p->t = nextopt(&p->a); |
329 | if (p->t == O_EOF) |
330 | p->a = eof; |
331 | } |
332 | |
333 | static void p_factor(p_ctx *p, node **nn) |
334 | { |
335 | node_un *n; |
336 | if (p->t == O_LPAREN) { |
337 | p_next(p); |
338 | p_expr(p, nn); |
339 | if (p->t != O_RPAREN) |
1d2d1062 |
340 | die("syntax error near `%s': missing `)'", *p->a); |
6e403221 |
341 | p_next(p); |
342 | } else if (p->t == O_NOT) { |
343 | n = xmalloc(sizeof(node_un)); |
344 | n->n.func = n_not; |
345 | *nn = &n->n; |
346 | p_next(p); |
347 | p_factor(p, &n->arg); |
348 | } else { |
349 | switch (p->t) { |
350 | case O_ANAG: *nn = anagram(p->a + 1); break; |
351 | case O_SUBG: *nn = subgram(p->a + 1); break; |
352 | case O_WILD: *nn = wildcard(p->a + 1); break; |
353 | case O_TRACK: *nn = trackword(p->a + 1); break; |
a10122de |
354 | #ifdef HAVE_REGCOMP |
355 | case O_REGEXP: *nn = regexp(p->a + 1); break; |
356 | #endif |
6e403221 |
357 | default: die("syntax error near `%s': unexpected token", *p->a); |
358 | } |
359 | p_next(p); |
360 | } |
361 | } |
362 | |
363 | static void p_term(p_ctx *p, node **nn) |
364 | { |
365 | node_bin *n; |
366 | for (;;) { |
367 | p_factor(p, nn); |
368 | switch (p->t) { |
369 | case O_AND: |
370 | p_next(p); |
371 | default: |
372 | break; |
6e403221 |
373 | case O_RPAREN: |
374 | case O_OR: |
375 | case O_EOF: |
376 | return; |
377 | } |
378 | n = xmalloc(sizeof(node_bin)); |
379 | n->left = *nn; |
380 | n->n.func = n_and; |
381 | *nn = &n->n; |
382 | nn = &n->right; |
383 | } |
384 | } |
385 | |
386 | static void p_expr(p_ctx *p, node **nn) |
387 | { |
388 | node_bin *n; |
389 | for (;;) { |
390 | p_term(p, nn); |
391 | if (p->t != O_OR) |
392 | break; |
393 | p_next(p); |
394 | n = xmalloc(sizeof(node_bin)); |
395 | n->left = *nn; |
396 | n->n.func = n_or; |
397 | *nn = &n->n; |
398 | nn = &n->right; |
399 | } |
400 | } |
401 | |
402 | /* --- @p_argv@ --- * |
403 | * |
404 | * Arguments: @int argc@ = number of command-line arguments |
405 | * @const char *const argv[]@ = vectoor of arguments |
406 | * |
407 | * Returns: A compiled node, parsed from the arguments. |
408 | * |
409 | * Use: Does the donkey-work of parsing a command-line. |
410 | */ |
411 | |
412 | static node *p_argv(int argc, const char *const argv[]) |
413 | { |
414 | p_ctx p; |
415 | node *n; |
416 | |
417 | av = argv; |
418 | ac = argc; |
419 | ai = 1; |
420 | p_next(&p); |
2668675c |
421 | if (p.t == O_EOF) { |
422 | usage(stderr); |
423 | pquis(stderr, "(Run `$ --help' for more detail.)\n"); |
424 | exit(EXIT_FAILURE); |
425 | } |
6e403221 |
426 | p_expr(&p, &n); |
427 | if (p.t != O_EOF) { |
428 | die("syntax error near `%s': rubbish at end of line (too many `)'s?)", |
429 | *p.a); |
430 | } |
431 | return (n); |
432 | } |
433 | |
434 | /*----- Main code ---------------------------------------------------------*/ |
435 | |
436 | /* --- @main@ --- * |
437 | * |
438 | * Arguments: @int argc@ = number of command-line arguments |
439 | * @char *argv[]@ = vector of argument words |
440 | * |
441 | * Returns: Zero on success, nonzero on failure. |
442 | * |
443 | * Use: Picks entries from a word list which match particular |
444 | * expressions. This might be of assistance to word-game types. |
445 | */ |
446 | |
447 | int main(int argc, char *argv[]) |
448 | { |
449 | node *n; |
450 | FILE *fp; |
451 | dstr d = DSTR_INIT; |
452 | char *p, *q, *l; |
453 | |
454 | ego(argv[0]); |
455 | n = p_argv(argc, (const char *const *)argv); |
456 | |
457 | if ((fp = fopen(file, "r")) == 0) |
458 | die("error opening `%s': %s", file, strerror(errno)); |
459 | for (;;) { |
460 | dstr_reset(&d); |
461 | if (dstr_putline(&d, fp) < 0) |
462 | break; |
463 | l = d.buf + d.len; |
464 | for (p = q = d.buf; p < l; p++) { |
465 | if (!isalnum((unsigned char)*p)) |
466 | continue; |
467 | *q++ = tolower((unsigned char)*p); |
468 | } |
469 | *q = 0; |
470 | d.len = q - d.buf; |
471 | if (n->func(n, d.buf, d.len)) { |
472 | fwrite(d.buf, 1, d.len, stdout); |
473 | fputc('\n', stdout); |
474 | } |
475 | } |
476 | if (!feof(fp)) |
477 | die("error reading `%s': %s", file, strerror(errno)); |
478 | fclose(fp); |
479 | return (0); |
480 | } |
481 | |
482 | /*----- That's all, folks -------------------------------------------------*/ |